A Deep Incremental Boltzmann Machine for Modeling Context in Robots

TL;DR

This paper introduces an incremental deep Boltzmann machine that adaptively models context in robotic environments, improving scene classification by dynamically expanding its structure.

Contribution

It presents a novel incremental deep model extending Restricted Boltzmann Machines that automatically adapts its structure for better context modeling in robots.

Findings

Converges to accurate scene contexts.

Performs better or on par with existing models.

Effectively extends structure as needed.

Abstract

Context is an essential capability for robots that are to be as adaptive as possible in challenging environments. Although there are many context modeling efforts, they assume a fixed structure and number of contexts. In this paper, we propose an incremental deep model that extends Restricted Boltzmann Machines. Our model gets one scene at a time, and gradually extends the contextual model when necessary, either by adding a new context or a new context layer to form a hierarchy. We show on a scene classification benchmark that our method converges to a good estimate of the contexts of the scenes, and performs better or on-par on several tasks compared to other incremental models or non-incremental models.